Build option to disable 64 Disk Drive support#649
Conversation
|
Ah, sorry forgot about the Visual C++ files:
I think those explain the AppVeyor build fail. Not familiar with configuring those and no way to test locally (no Windows box here) so will need to read up. |
0ad88dc to
6d116fd
Compare
|
It seemed like there was a problem with the checks for desktop builds because the Makefile wasn't setting the pre-processor macro, but now I'm not so sure.
I hope this feature is acceptable. |
|
I'm okay with this capability in general (being able to disable 64DD support at build time) but I have a problem with this implementation. There are 50 new #ifdef macro blocks in this merge request (that's a lot), and the function interfaces (internal API) are different depending upon whether or not this macro is defined. There must be a simpler way to achieve this with fewer #ifdefs, like setting the dd_controller pointer to NULL if it's not supported, and keeping the existing internal API. |
|
I think it would be a lot better to figure out why the new dynarec crashes with 64DD support, there shouldn't really be a reason to remove DD support like this |
|
Also, Android does not appear to have this problem. I agree with @loganmc10, if @hissingshark could provide a backtrace to the core dump, I'm sure the issue could be figured out. |
There is a longstanding issue since the addition of the 64 Disk Drive support. This is causing some ARM devices to segfault at emulator start.
This commit adds a build option
64DISKDRIVE.It defaults to
64DISKDRIVE=1as I expect most devices are unaffected.Those platforms vulnerable to it can use
64DISKDRIVE=0.Many Thanks